Power quality in hospitals describes the extent to which the voltage, current and frequency supplied to an installation retain the characteristics required for connected loads to operate as intended. A hospital may maintain power continuity and still experience harmonics, voltage dips, transients or imbalances capable of increasing losses, overheating components or disrupting control systems.
The useful question is not whether electricity is present, but whether the waveform and its behaviour over time are compatible with the installation, the connected loads and the criticality of the area. A rigorous answer requires measurements taken over a representative period, correlation between events and actual operating conditions, and subsequent verification of any corrective action.
Electricity leaves traces even when the supply never fails. They appear in a distorted waveform, in the abnormal heat of a neutral conductor or in a restart that no one can reproduce. Measuring well means following those traces from symptom to source, then returning to the same point to prove that the correction has truly worked.
What power quality means in hospitals
Power quality is not a single value. It is the combined result of several quantities that change over time and must be interpreted at the point where they are measured. A deviation observed at the incoming supply may originate outside the facility; one recorded only downstream of a distribution board may be associated with a particular load or with the internal distribution system itself.
The Department of Veterans Affairs Electrical Design Manual requires voltage and current harmonic distortion to be estimated at buses serving a high proportion of non-linear loads and, where appropriate, mitigation measures to be justified. The principle matters: first characterise the problem; then decide whether it needs to be corrected.
| Disturbance | What is observed | What should be investigated |
|---|---|---|
| Harmonics | Periodic distortion of voltage or current | Non-linear loads, variable-speed drives, switched-mode power supplies and resonances |
| Voltage dip | Temporary reduction in RMS voltage | Starts, faults, transfers and grid events |
| Transient | Rapid, short-duration variation | Switching operations, discharges and inadequate suppression |
| Imbalance | Differences between phases | Load distribution, connections and changes in demand |
| Frequency variation | Deviation from nominal frequency | Supply source, local generation and changeovers |
Why electrical harmonics need context
A linear load tends to draw a current whose shape follows the applied voltage. Non-linear loads draw current non-sinusoidally and may introduce components at multiples of the fundamental frequency. The result is commonly expressed as total harmonic distortion, or THD, but the percentage alone explains neither the source nor the impact.
The U.S. Department of Energy guide notes that harmonics can harm transformers and motors and interfere with sensitive control equipment. It also warns that power factor correction and resonances within the installation can alter the behaviour of the problem. A filter should therefore never be specified on the basis of a single reading.
Voltage THD and current THD answer different questions
Current THD helps characterise how a load, or group of loads, draws power. Voltage THD shows the extent to which that demand, together with the network impedance, is distorting the voltage at the measurement point. Comparing both records, as well as the spectrum by harmonic order, provides considerably more insight than a single aggregate indicator.
Instantaneous load also matters. A high percentage measured at very low current may have a different significance from persistent distortion under peak operating conditions. The survey should record power, RMS current, power factor, imbalance and events—not THD alone.
The neutral conductor and temperature provide complementary evidence
In three-phase distributions containing numerous single-phase non-linear loads, certain harmonics may accumulate in the neutral conductor. A technical investigation should compare phase and neutral currents, verify load distribution and inspect temperatures at connections, conductors and transformers. Thermography can reveal symptoms, but it cannot replace electrical analysis or prove the cause on its own.
Voltage dips, transients and imbalances: disturbances hidden by averages
A ten-minute average may appear normal while concealing an event lasting only a few cycles. An analyser used to investigate restarts, communication losses or intermittent alarms must therefore capture events with a resolution and trigger configuration suited to the symptom. The recording period should also encompass different shifts, equipment starts, supply transfers and changes in activity.
Voltage dips
A voltage dip may coincide with the starting of a large load, a short circuit elsewhere, a supply transfer or a grid event. If it appears simultaneously at the incoming supply and downstream, the investigation will follow a different path from that of an event recorded only at a secondary distribution board. Comparing time-aligned measurements at different points prevents the equipment that displayed the failure from being mistaken automatically for its cause.
Transients
Transients may be associated with switching operations, capacitor-bank switching, interruption of inductive loads or external phenomena. Their duration demands suitable instrumentation and capture settings. A conventional multimeter may confirm steady-state values, but it will not necessarily reconstruct a fast event.
Phase imbalance
Imbalance can increase current, raise losses and affect the behaviour of three-phase loads. Measurements should include voltage and current on every phase and be repeated under comparable operating conditions. If the distribution changes throughout the day, the solution may involve redistributing loads or reviewing the installation before adding corrective equipment.
How to design a useful measurement survey
Power quality is best diagnosed as a process rather than through an isolated inspection. Before connecting an instrument, formulate a testable hypothesis: what happens, where it occurs, how often it occurs and what changed before it first appeared. This definition prevents large volumes of data from being collected without a technical question to answer.
- Define the symptom. Identify the equipment, area, time, duration, alarm and observed consequence.
- Select the measurement points. Where safe and technically feasible, compare the incoming supply, main busbar and a point close to the load.
- Configure the capture. Select intervals, thresholds and waveform recording settings appropriate to the event.
- Record the context. Log transfers, starts, tests, building work, load changes and incidents.
- Correlate the evidence. Align the electrical data chronologically with operating records.
- Correct and repeat. Apply a justified measure and verify it under comparable conditions.
Where to measure
The measurement point changes the meaning of the result. The incoming supply reveals interaction with the grid; a main busbar shows aggregate behaviour; and a point close to the load captures the conditions actually received by the equipment. A staged survey is usually more useful than installing analysers without a comparison architecture.
The work must be carried out by qualified personnel using appropriate safety procedures, instruments suitable for the measurement category and a prior assessment of access conditions. This article does not replace an electrical design, a safe system of work or the manufacturer’s instructions.
How measurements relate to critical medical areas
In operating theatres and intervention rooms, IEC 60364-7-710:2021 establishes particular requirements for electrical installations in medical locations, covering the protection of patients and medical staff. Power quality complements these requirements; it neither replaces them nor allows compliance to be inferred from a partial measurement.
IT power supply systems maintain continuity after a first insulation fault in their intended applications, while IT insulation monitoring enables the corresponding deterioration to be detected. Neither function is equivalent to a general-purpose power quality analyser: they are separate layers that must work together.
Similarly, an uninterruptible power supply system can provide continuity during certain disturbances, but its behaviour depends on its topology, configuration, load, autonomy and installation conditions. No UPS should be assumed to eliminate every harmonic or transient.
From data to corrective action
A correction must address a demonstrated mechanism. If the source is traced to a non-linear load, the options may include architectural changes, line reactors, active or passive filters, load redistribution or equipment with better input characteristics. If the problem is a voltage dip, the answer may lie in coordination, backup supply, starting arrangements or the grid. There is no universal solution.
| Evidence | Hypothesis to test | Minimum validation |
|---|---|---|
| High current THD on one branch | Dominant non-linear load | Measure before and after operating that load |
| Voltage THD at several distribution boards | Propagated distortion or common impedance | Compare the incoming supply, busbar and branches |
| Simultaneous dip at the incoming supply and load | Upstream event | Correlate timestamps and grid records |
| Elevated neutral temperature | Harmonic current or poor connection | Measure current, spectrum, connection torque and temperature |
| Restart with no captured event | Unsuitable trigger threshold or resolution | Reconfigure the capture and repeat |
After any intervention, the survey should be repeated using the same location, instrumentation and load conditions wherever possible. Improvement is demonstrated by comparing records—not merely by the symptom failing to reappear for a few hours.
When a disturbance does not interrupt the supply
Not every electrical anomaly ends in a disconnection. Protection coordination in hospitals determines which part of the installation must be isolated when a fault occurs; power quality analysis, by contrast, reveals disturbances capable of degrading performance without operating any protective device. The two disciplines address different moments in the same problem: containing the fault when it appears and recognising the signs that may precede it.
Those signs become genuinely useful when they are observed over time. A sustained change in THD, imbalance or temperature can inform predictive maintenance in hospitals and critical areas, provided that measurements are comparable and thresholds reflect the installation’s actual behaviour. A single value describes a moment; a trend reveals whether the system itself is changing.
When the source is unclear or several layers of the distribution system are involved, hospital electrical safety consulting helps organise the evidence before a decision is made. The electrical diagram, connected loads, criticality of each area and event history must guide the diagnosis. Naming a disturbance is only the beginning; locating it and demonstrating its effect is what makes a technically sound correction possible.
Frequently asked questions about hospital power quality
What is power quality in hospitals?
It is the assessment of voltage, current, frequency and electrical disturbances in relation to the requirements of the installation and its loads. It encompasses harmonics, voltage dips, transients and imbalances, together with their duration, location and operational effect.
Does a high THD prove that a filter is required?
No. Voltage and current THD must first be distinguished, and the load, harmonic spectrum, source, network impedance and possible resonances must be evaluated. A filter can be justified only after the underlying mechanism has been characterised.
Does a UPS eliminate every electrical disturbance?
Not necessarily. Its response depends on its topology, configuration, load and specifications. Performance must be verified against the particular disturbances and within the installation’s complete electrical architecture.
How long should a measurement survey last?
It should cover representative conditions, including different shifts, maximum and minimum load, transfers and events related to the symptom. There is no universal duration; the objective is to capture enough operating cycles to support a valid comparison.
Where should the power quality analyser be installed?
At points selected to test specific hypotheses: the incoming supply, main busbar and a location close to the load, depending on risk and accessibility. Simultaneous or time-aligned comparisons help distinguish external sources from internal ones.
How is a corrective measure validated?
By repeating the measurements under comparable conditions, using the same indicators and points, and confirming both a reduction in the disturbance and the absence of adverse effects elsewhere in the installation.
Turn a disturbance into a verifiable decision
If you need to characterise an installation before adding loads, investigate incidents or define corrective measures, consult ETKHO’s electrical safety advising service.